41

Chapter 3

ACTING AS WE FEELWhen and How Attitudes Guide Behavior

RUSSELL H. FAZIO DAVID R. ROSKOS-EWOLDSENThe Ohio State University University of Alabama

Consider each of the following statements. Do you believe the statement to be true or false?

1. College students who disapprove of cheating do not cheat on tests; it is only the students whoview cheating as acceptable who do cheat.

2. When segregation was still legal, hotel and restaurant owners with racial stereotypes towardChinese people would not serve them food or allow them to stay at their establishments.

3. How well people like their jobs is predictive of people’s job attendance. Those who like theirjobs are less likely to miss a day of work.

4. During the 1970s, people who felt that the energy crisis was a significant problem used lessenergy than did those who did not really believe that there was a crisis.

5. Regardless of whether an employer makes a snap judgment or deliberates extensively about ahiring decision, if the employer has a negative attitude toward working women, a female candi-date will not be hired.

All of these commonsense statements assume that people’s attitudes influence their actionsand decisions. In fact, as we will see in this chapter, none of these five statements is correct.The basic finding of decades of research is that sometimes people act in accordance with theirattitudes, and other times they act in ways that are quite inconsistent with their attitudes.

In this chapter, we address three fundamental questions regarding the attitude–behaviorrelation (Zanna & Fazio, 1982). First, is there a relation? That is, do attitudes influence behavior?Second, when is such a relation to be expected? In other words, what variables determine thedegree to which attitudes might influence behavior? To the extent that attitudes do predict behav-ior, this question concerns the identification of other factors that play a role in this relationship.

03-Brock.qxd 12/1/2004 10:51 AM Page 41

Finally, how do attitudes guide behavior? By what psychological processes do attitudes exertthese influences? If we are to understand the relation between attitudes and behavior, we need todevelop models and theories of the psychological processes that link attitudes to behavior.Furthermore, understanding the psychological processes underlying the attitude–behavior rela-tionship has many practical implications. For example, we can design better health campaignsto counter unhealthy behaviors if we understand how attitudes relate to behavior (Roskos-Ewoldsen, Yu, & Rhodes, 2004).

IS THERE A RELATION BETWEENATTITUDES AND BEHAVIOR?

For a number of decades, the field of social psychology has had reason to question theintuitively reasonable assumption that people act on the basis of their attitudes. During theearly 1930s, LaPiere (1934) conducted what has become probably the most widely citedstudy of the attitude–behavior relation. While traveling across the western United States in thecompany of a Chinese couple, LaPiere stopped at more than 200 hotels and restaurants. TheChinese couple was refused service at only one establishment. Some 6 months later, LaPierewrote to the proprietor of each of the hotels and restaurants and asked whether the establish-ment served Chinese guests. Surprisingly, 92% of those who responded indicated that theydid not accommodate Chinese guests. Thus, there was a startling inconsistency between theattitude responses to LaPiere’s letter and the actual behavior toward the Chinese couple withwhom LaPiere had traveled. A very similar study concerning an African American guest,instead of Chinese guests, also observed much discrepancy between people’s reports of theirattitudes and their actual behavior (Kutner, Wilkins, & Yarrow, 1952).

Although these findings seem to indicate a lack of correspondence between attitudes andbehavior, the relevance of these classic studies to the issue of attitude–behavior consistencyhas been questioned. For example, the point has been raised that the persons who waited onthe Chinese guests in LaPiere’s (1934) study or the African American guest in Kutner and col-leagues’ (1952) study might not have been the same persons who responded to the attitudequestion (Ajzen, Darroch, Fishbein, & Hornik, 1970; Dillehay, 1973). In addition, it can beargued that the specific individuals who were admitted to the establishments in these studieswere not representative of what came to the proprietors’ minds when asked in an abstractmailing about admitting Chinese or African American individuals (Lord, Lepper, & Mackie,1984). That is, the proprietors may have imagined slovenly unappealing persons whenresponding to the attitude question, in contrast to the pleasant appearance of the specific indi-viduals who were admitted.1

However, these studies are by no means the only ones to challenge the assumption thatpeople typically behave consistently with their attitudes. For example, Corey (1937) exam-ined the relationship between students’ attitudes toward cheating and their actual cheatingbehavior. The students took a series of true/false examinations, which they self-scored at alater class meeting. The students did not know, however, that the instructor had scored theexams during the interim period. Thus, the difference between the scores that studentsassigned to themselves and the scores that the instructor assigned served as the measure ofstudents’ cheating behavior. The correlation between the students’ attitudes toward cheatingand actual cheating was essentially zero. Attitudes toward cheating did not in the least bit

42— P E R S U A S I O N

03-Brock.qxd 12/1/2004 10:51 AM Page 42

predict the actual cheating behavior. Instead, cheating was related to test performance; themore poorly students had done on the exam, the more likely students were to cheat in scoringthe exam (as Calvin demonstrates in Figure 3.1).

Corey’s (1937) findings are not unusual. Indeed, in a highly influential article, Wicker(1969) reviewed 31 investigations of the attitude–behavior relation and concluded,

Taken as a whole, these studies suggest that it is considerably more likely that attitudes willbe unrelated or only slightly related to overt behaviors than that attitudes will be closelyrelated to actions. . . . Correlation coefficients relating the two kinds of responses arerarely above .30 and often are near zero. (p. 65)

Wicker’s review, along with others (e.g., Deutscher, 1973), led to considerable skepticism—sufficiently so that some suggested, “It may be desirable to abandon the attitude concept”(Wicker, 1971, p. 29).

Nonetheless, this skepticism does not appear to have been fully warranted. Although it can-not be denied that a large number of studies suggest that attitudes do not influence behavior,sometimes attitudes do predict behavior. For example, studies of voting behavior consistentlyhave indicated a substantial relation between preelection attitudes and voting. Basically, peoplevote for the candidates they like. Kelley and Mirer (1974) analyzed data concerning the fourpresidential elections from 1952 to 1964 and found that voting behavior could be predictedaccurately from preelection attitudes for 85% of the respondents.

Likewise, in a study on organ transplants, participants initially indicated their attitudestoward organ transplantation (Goodmonson & Glaudin, 1971). Later, the experimenters made aseries of successively more difficult and more committing requests of the respondents, fromrequesting that they schedule an appointment to be interviewed about organ transplants, torequesting that they participate in the interview, to requesting that they actually sign a legal doc-ument providing posthumous organ donation. The number of behavioral steps that the partici-pants took toward this final goal served as the index of behavior. The correlation betweenattitudes and behavior was .58. In a study of homeowners’ actual energy consumption, a negli-gible correlation was found between the perceived severity of the energy crisis and energy use,

Acting as We Feel— 43

Figure 3.1 Positive Attitude Toward Cheating, or Cheating Motivated by Poor Performance?

SOURCE: Calvin and Hobbes © 1988 Watterson. Reprinted with permission of Universal Press Syndicate. All rightsreserved.

03-Brock.qxd 12/1/2004 10:51 AM Page 43

but a correlation of .65 was observed between energy use and homeowners’ attitudes regardingthe necessity of air-conditioning in maintaining their health and comfort (Seligman et al., 1979).

Even this brief sampling of positive findings indicates that attitude–behavior correlations canand sometimes do exceed the .30 ceiling claimed by Wicker (1969). Consideration of positivefindings of this sort has led to a far more optimistic outlook about the usefulness of attitudes inpredicting behavior (Calder & Ross, 1973; Fazio & Zanna, 1981; Schuman & Johnson, 1976;Zanna & Fazio, 1982). The blanket statement that attitudes have little to do with behavior isoften contradicted by studies in the literature. Research has revealed everything from findingsof no relation whatsoever (e.g., Corey’s [1937] study of cheating behavior) to the nearly perfectrelation observed in the context of voting behavior.

Thus, the answer to the question “Is there a relation between attitudes and behavior?” is aresounding “sometimes.” Given the range of findings, it becomes apparent that the questionof attitude–behavior consistency has to be approached differently:

Rather than asking whether attitudes relate to behavior, we have to ask, “Under what con-ditions do what kinds of attitudes of what kinds of individuals predict what kinds of behav-ior?” We need to treat the strength of the attitude–behavior relation as we would treat anyother dependent variable and determine what factors affect it. (Fazio & Zanna, 1981,p. 165)

WHEN DO ATTITUDES GUIDE BEHAVIOR?

This question calls for identifying factors that determine whether the relation between atti-tudes and behavior will be relatively strong or weak. Such factors are typically referred to asmoderating variables because they moderate the relation between attitudes and behavior. Aswas hinted earlier, moderators of the attitude–behavior relation include qualities of the behav-ior, qualities of the person, qualities of the situation in which the behavior is exhibited, andqualities of the attitude itself. We review briefly the evidence regarding each of these classesof potential moderating variables.

Qualities of the Behavior

The behaviors that a social psychologist might be interested in predicting from knowledgeof a person’s attitudes can range from the very specific (e.g., will the person attend churchservices this week?) to the very general (e.g., how many religious behaviors will the personperform over the next month?). In a highly influential analysis, Ajzen and Fishbein (1977)noted the importance of measuring attitudes and behavior at equivalent levels of specificity.A specific behavior is best predicted by an attitudinal question that is equivalently specific tothe action in question, the target of the action, the context in which the action is performed,and the time of the action (e.g., “How do you feel about attending church this Sunday?”). Ina study conducted prior to the mandated use of lead-free gasoline, the actual purchase of lead-free gas was better predicted by questions asking specifically about buying lead-free gas thanby questions assessing more general attitudes toward ecology (Heberlein & Black, 1976).

In contrast, a general pattern of behavior is best predicted by a general attitude measure.In one study, participants’ global attitude toward “being religious” was used to predict the

44— P E R S U A S I O N

03-Brock.qxd 12/1/2004 10:51 AM Page 44

likelihood that they performed each of 100 specific religious behaviors (e.g., praying beforeor after meals, donating money to a religious institution) and a general measure of performingreligious behaviors that was a composite measure of the 100 specific religious behaviors(Fishbein & Ajzen, 1974). The correlation between the general attitude toward religion andany specific single action was a mere .15 on average. In contrast, the correlation between atti-tude and the general behavior pattern (i.e., the number of religious actions performed) was.71. In their review of the literature, Ajzen and Fishbein (1977) noted that studies thatemployed attitude and behavior measures that were equally specific typically found higherattitude–behavior correlations than did studies in which one of the two measures was morespecific than the other. Thus, the degree of match between the attitude and the behavior wewish to predict affects the strength of the attitude–behavior relation that will be observed.

Qualities of the Person

In addition, some kinds of people typically display greater attitude–behavior consistencythan do others. In general, two classes of individuals have been considered: those who areaware of and guided by their internal feelings and those who tend to rely heavily on cues inthe situation to decide how to behave. In general, people who are aware of their feelingsdisplay greater attitude–behavior consistency than do people who rely on situational cues.

Obviously, this is a very rough distinction. Any given behavior of an individual can beguided both by the individual’s internal feelings and by external cues. Yet a number of per-sonality scales have been developed and used successfully to assess whether a given persontends to rely more heavily on one type of cue or the other. Although some important differ-ences exist among the personality traits that have been explored as possible moderators of theattitude–behavior relation, each relates to this general distinction. Level of moral reasoninghas been found to affect the relation between attitudes and behavior (Rholes & Bailey, 1983).More advanced moral reasoning is characterized by principled, morally responsible thoughtbased on people’s own general principles of moral action. Lower levels of reasoning focus onthe general positive or negative consequences of a particular action or on a feeling of beingbound by social or legal rules. Individuals who depend on their own feelings and principlesto make moral judgments act much more consistently with their attitudes toward moral issuesthan do people who rely on external standards to determine what is moral.

The personality dimension that has received the greatest attention in the context of theattitude–behavior issue is self-monitoring (Snyder, 1987). Individuals who score low on theself-monitoring scale claim to be guided by dispositions (i.e., their inner feelings). They agreewith statements such as “My behavior is usually an expression of my true inner feelings, atti-tudes, and beliefs.” In contrast, individuals who score high on the self-monitoring scale viewtheir behavior as stemming typically from a pragmatic concern with what is appropriate ineach situation. They agree with statements such as “In different situations and with differentpeople, I often act like very different persons.” Thus, these individuals are said to monitor theimpression that they make on other people and adjust that impression to fit with others’ expec-tations. A number of studies have indicated that low self-monitors behave more consistentlywith their attitudes than do high self-monitors (Ajzen, Timko, & White, 1982; Snyder &Kendzierski, 1982; Snyder & Swann, 1976; Zanna, Olson, & Fazio, 1980). For example, in astudy on the relation between attitudes toward affirmative action and judgments of liabilityin a simulated sex discrimination case, the correlation between participants’ attitudes and

Acting as We Feel— 45

03-Brock.qxd 12/1/2004 10:51 AM Page 45

their judgments of liability was .42 among low self-monitors. Among high self-monitors, thecorrelation was a negligible .03 (Snyder & Swann, 1976).

To reiterate, both of these personality types vary on the extent to which the individualpays attention to or is influenced by his or her internal feelings. As one would expect,people who focus on themselves tend to act more consistently with their attitudes (e.g.,people with high moral reasoning and/or low self-monitoring). On the other hand, peoplewho are guided more by the environment or other external factors often do not act in amanner that is consistent with their attitudes (e.g., people with low moral reasoning and/orhigh self-monitoring).

Qualities of the Situation

A number of situational variables also affect the strength of the attitude–behavior relation.These include normative factors and time pressure to reach a decision.

The Effect of Norms

Norms, or beliefs about how one should or is expected to behave in a given situation, canexert a powerful influence on behavior. Fishbein and Ajzen (1975) proposed a model thatviews norms as having a major influence on behavior (we review this model more fully in alater section). Much evidence has been found in support of this view (for a review, see Ajzen& Fishbein, 1980). People often behave as they believe others expect them to behave.

Norms can constrain an individual’s behavior to the point where it is unlikely that theperson will display behavior consistent with his or her attitudes (Figure 3.2). Indeed, anorm may be so strong and so universally held that virtually everyone in that situationbehaves the same regardless of his or her attitude. For example, you might wish that some-one were dead, but you would very rarely act on this attitude. Hence, attitude–behaviorconsistency is low. Consider also the relation between job satisfaction and work atten-dance. At first glance, one might expect people who like their jobs to be less likely to missa day of work. Yet the normative pressure (in addition to the potential financial pressure)to attend work every day is strong. Thus, with the exception of days when they are ill,people generally go to work every day, even people who do not like their jobs. Indeed,studies of job satisfaction have found little relation between attitudes toward one’s job andabsenteeism (e.g., Vroom, 1962).

However, consider what might happen on a day when an unforeseen event does free indi-viduals from their sense of obligation to attend work. A severe snowstorm strikes, makingtravel very difficult and also making it clear that not everyone will get to work that day.Precisely such a situation was studied in a company on the day following a major snowfall inChicago (Smith, 1977). The attendance rate that day was approximately 70%, and work atti-tudes did predict attendance. Averaging across six different attitude measures, the correlationbetween work attitudes and attendance was .46. In contrast, a comparison sample from thesame company’s office in New York, where no snowstorm had occurred on this particular day,had an attendance rate of approximately 96% and revealed an average correlation betweenattitudes and behavior of only .08.

46— P E R S U A S I O N

03-Brock.qxd 12/1/2004 10:51 AM Page 46

Time Pressure

Individuals are more likely to base their decisions on their attitudes when they are undertime pressure because their attitudes provide a heuristic for making quick decisions (for areview, see Jamieson & Zanna, 1989). It appears that time pressure pushes people away froma careful examination of the available information and toward a reliance on their preexistingattitudes. For example, in one study, participants were asked to consider job applications fromboth male and female job candidates (Bechtold, Naccarato, & Zanna, 1986). When there wasno time pressure, and so participants could consider all of the details carefully, their person-nel decisions were unrelated to their attitudes toward working women. That is, participantswhose earlier reported attitudes indicated some prejudice against women were just as likelyto recommend hiring a female candidate as were those who did not hold such prejudiced

Acting as We Feel— 47

Figure 3.2 Strong Influence of Norms on Behavior Regardless of Attitudes

SOURCE: Gahan Wilson. Copyright © 1986. Reprinted with permission from The New Yorker Magazine, Inc.

03-Brock.qxd 12/1/2004 10:51 AM Page 47

attitudes. In striking contrast, when participants were under time pressure to make a hiringrecommendation, an attitude–behavior relation was apparent. Participants who were preju-diced against women were less likely to recommend hiring a female candidate.

The latter study is an interesting one because it points out that, from a societal perspective,there are some instances when attitude–behavior consistency is not desirable. In this instance,acting in accordance with an attitude leads to discrimination against certain groups within oursociety (Amodio & Devine, chap. 11, this volume). Jamieson and Zanna (1985) also foundthat, in a simulated sex discrimination lawsuit and a simulated trial involving a mandatorydeath penalty, if the defendants were judged to be guilty, attitudes toward affirmative actionwere predictive of judgments in the sex discrimination lawsuit and attitudes toward capitalpunishment were predictive of judgments of guilt in the criminal trial—provided that partic-ipants were under time pressure to read the case material and reach a decision. Thus, in threesituations where one would hope that individuals would consider the details of a case objec-tively and be free of the bias of their attitudes, participants were able to do so when they wereallowed to examine the case material at their own pace. However, when they were under timepressure, participants were strongly biased by their existing attitudes.

Qualities of the Attitude

Some kinds of attitudes appear to be stronger than others (Petty & Krosnick, 1995). In thiscontext, the word stronger is not used in the sense of the attitude being more extreme. Instead,stronger refers to the apparent influence that the attitude has on the individual’s behavior.In fact, in all of the research that is summarized in this section, groups of participants withdifferent degrees of attitude strength were compared, but the distributions of attitude scores(i.e., the extremity of attitudes) in the various groups were equivalent to one another.

The Role of Direct Experience

One attitudinal quality that has been investigated extensively is the manner of attitude for-mation (for a review, see Fazio & Zanna, 1981). On the one hand is attitude formation throughdirect behavioral experience with the attitude object, and on the other hand is attitude forma-tion through indirect nonbehavioral experience with the attitude object. For example, a childmay form an attitude toward a toy by playing with the toy (direct experience) or on the basisof a friend’s or an advertisement’s description of the toy (indirect experience).

Attitudes based on direct experience have been found to be more predictive of later behav-ior than attitudes based on indirect experience. This was first shown in a study that tookadvantage of an actual event at Cornell University (Regan & Fazio, 1977). Because of a cam-pus housing shortage, many freshmen had spent the first few weeks of the academic year intemporary housing. Typically, these accommodations consisted of cots in the lounge of a dor-mitory. Those freshmen who were assigned to temporary quarters had much more direct expe-rience with the housing crisis than did the freshmen who were immediately assigned topermanent housing. Those freshmen who were assigned to permanent quarters, on the otherhand, had learned about (and formed their attitudes toward) the housing shortage only throughengaging in discussions with others and reading the frequent articles on the subject in thecampus newspaper. Thus, a naturally occurring event had created two groups that differed intheir manner of attitude formation. The two groups were compared in terms of the extent to

48— P E R S U A S I O N

03-Brock.qxd 12/1/2004 10:51 AM Page 48

which they displayed behavior that was consistent with their attitudes toward the housingcrisis (e.g., agreeing to sign a petition calling on the university to alleviate the shortage,obtaining the signatures of other students, writing letters to the university housing office).Attitude–behavior consistency was much greater among those who had been assigned totemporary housing (the direct experience group) than among those who had been assigned topermanent housing (the indirect experience group). This was true even though the two groups,on the average, had equally negative attitudes toward the housing shortage.

To further establish the role of direct experience in the attitude–behavior relationship, anexperiment was conducted where the manner of attitude formation was manipulated in thelaboratory (Regan & Fazio, 1977). Participants were introduced to a set of five intellectualpuzzles in one of two ways. Some participants were presented previously solved examples ofeach puzzle and listened to the experimenter describe the type of puzzle and the specificexample and solution (indirect experience condition). The remaining participants were givenan opportunity to actually work the same example puzzles, thereby forming their attitudesthrough direct behavioral experience. After attitudes toward each of the five types of puzzleswere assessed, all individuals participated in a free play situation. That is, they were givennumerous samples of each puzzle type and instructed to play with any that they so desired.On average, the relation between a given participant’s attitude toward a puzzle and the amountof free play behavior with the puzzle was greater in the direct experience condition than inthe indirect experience condition (see also Fazio & Zanna, 1978).

Attitude Accessibility

One thing that differentiates attitudes based on direct experience from those based on indi-rect experience is how accessible the attitudes are from memory. Accessibility in this senserefers to how easily attitudes come to mind. Some attitudes come to mind without any con-scious effort on people’s part. When people see a cockroach, the “Yuck!” response probablycomes to mind immediately. This attitude would be highly accessible from memory. Butsometimes people have to deliberate quite extensively about what their attitudes toward someobject are. If you are asked which of several restaurants is the best Tibetan restaurant, youmight have to think extensively about which one you like the best. This attitude would not beat all accessible from memory.

As these examples illustrate, one way in which to measure how accessible an attitude isfrom memory is by how long it takes people to answer whether they like or dislike something.Attitudes based on direct experience tend to be more accessible (e.g., can be expressed morequickly) from memory (Fazio, Chen, McDonel, & Sherman, 1982). Participants who haddirect experience with puzzles were able to respond more quickly to inquiries about theirattitudes toward the puzzles than were participants who had indirect experience. As will bediscussed later, attitude accessibility plays a major role in the attitude–behavior relation.

Subsequent research on attitude accessibility has explored the functional value of such atti-tudes (for an extensive review, see Fazio, 2000). That is, what do accessible attitudes do forthe individual? How do they help the individual to navigate the day-to-day world? Researchhas found that accessible attitudes ease decision making (Blascovich et al., 1993; Fazio,Blascovich, & Driscoll, 1992). Imagine what it would be like if every time you went into aBaskin-Robbins, you had to decide which flavor of ice cream you wanted by reviewing theentire list of offerings and considering the relative merits of each type of ice cream. You would

Acting as We Feel— 49

03-Brock.qxd 12/1/2004 10:51 AM Page 49

probably take a long time to make the decision, and the decision would probably be stressful(enhanced all the more by the stress you would feel from the people behind you in line impa-tiently waiting for you to make up your mind). However, if the fact that you really like twoflavors—mint chocolate chip and pistachio almond—readily comes to mind, the decisionbecomes much easier (especially if you order a scoop of each flavor). As the research on atti-tudes and time pressure demonstrated, attitudes can serve as useful heuristics for decisionmaking. Because accessible attitudes come to mind readily, they make the decision-makingprocess that much easier.

Research has demonstrated that influence of accessible attitudes can have important effectson people’s well-being. In a study of college freshmen, the accessibility of the freshmanresearch participants’ attitudes toward academically relevant issues (e.g., possible majors,specific courses, academic activities such as studying at the library and pulling an all-nighter)was measured during their first 2 weeks at the university. The participants also completed var-ious inventories regarding their mental and physical health as well as a report of any stressorsthey were experiencing (Fazio & Powell, 1997). Two months later, participants completedthese measures again. The findings of this study clearly illustrated the utility of accessibleattitudes in handling the stress related to being a college freshman. For students starting theircollege careers in good health, having accessible attitudes toward the new experiences of col-lege buffered them from the negative effects of stress. These students were less likely to expe-rience negative effects of stress, such as physical illness and depression, if they had accessibleattitudes toward the new experiences found at college. Among students who started college inrelatively poor health, those with more accessible attitudes showed greater recovery overtime. Thus, there was considerable value to students knowing their likes and dislikes regard-ing the many new issues that they were encountering as they adjusted to college life.

Other research on attitude accessibility suggests that at least part of the benefits of accessi-ble attitudes may arise due to their influence on how people attend to and process informationin their day-to-day environment. For example, accessible attitudes influence what informationpeople attend to in their surroundings (Roskos-Ewoldsen & Fazio, 1992). Consequently, peopleare more likely to notice and pay attention to elements of their environment that are attitudinallyimportant to them. Likewise, accessible attitudes influence how people categorize and interpretinformation. One characteristic of social situations is that most stimuli are open to a multitudeof classifications. When interacting with a person, do you pay attention to that person’s race,gender, age, physical characteristics, and so forth? Indeed, the accessibility of your attitudestoward the various possible categorizations of an object influences how the object is categorized(Smith, Fazio, & Cejka, 1996). The late Mother Teresa could be categorized as a humanitarianor a Catholic. The accessibility of participants’ attitudes toward the categories of humanitarianand Catholic influenced how they categorized Mother Teresa. If research participants had moreaccessible attitudes toward Catholicism, she was more likely to be categorized as a Catholic.Likewise, research has demonstrated that as the accessibility of participants’ attitudes towardrace increased, the more likely they were to categorize novel people in terms of race (Fazio &Dunton, 1997).

Discussing the functional value of accessible attitudes implies that accessible attitudes per-form a number of useful functions for people, and indeed they do. However, there is a darkside to accessible attitudes as well. Accessible attitudes may be extremely difficult to change,with the upshot that people may be rather close-minded concerning topics toward which theyhave accessible attitudes. Because of the influences of accessible attitudes on what is attended

50— P E R S U A S I O N

03-Brock.qxd 12/1/2004 10:51 AM Page 50

to and how that information is categorized, accessible attitudes create a dynamic with theenvironment such that the accessible attitudes naturally reinforce their accessibility (Roskos-Ewoldsen, 1997). In addition, people actually display some difficulty in detecting changes inobjects toward which they have accessible attitudes (Fazio, Ledbetter, & Towles-Schwen,2000). In a series of three experiments, participants were shown photographs of variouspeople. Some participants were induced to rehearse their attitudes toward the photos and,hence, developed accessible attitudes. Those in a control condition saw the photos just asmany times but performed an attitude-irrelevant task whereby they were asked to estimate thetarget person’s height. During a second phase of the experiments, the participants judgedwhether each of a series of photos was the same as or different from the pictures they had seenearlier. Some of these photos had been morphed to various degrees with a picture of a differ-ent second person. When participants had rehearsed and developed accessible attitudestoward the persons in the original photograph, more change in the photo was necessary forthem to detect that the photo was different. The accessible attitudes apparently “blinded” theparticipants, detracting from their ability to detect small changes.

Summary

The findings we have reviewed in this section make it abundantly clear that attitudes dosometimes relate to behavior. Extreme pessimism regarding the value of attitudes as predic-tors of behavior is unwarranted. Furthermore, we now have a lengthy catalog of situational,personality, attitudinal, and behavioral qualities that appear to determine the strength of theattitude–behavior relation.

What is missing, however, is any sense of why these various factors exert their influence. Whydo only certain kinds of attitudes or certain kinds of situations promote attitude–behavior consis-tency? These concerns raise a very basic question regarding the attitude–behavior relation: Howdo attitudes guide behavior? That is, by what processes do attitudes influence behavior?

If we had an understanding of such processes, it would be far easier to understand whyonly certain kinds of attitudes or certain kinds of individuals in certain kinds of situationsseem to guide behavior. It is to this point that we now turn.

HOW DO ATTITUDES GUIDE BEHAVIOR?

Two different mechanisms by which attitudes can influence behavior are discussed in thissection. The major distinction between the two mechanisms centers on the extent to which thebehavior is thoughtfully planned in advance of its actual performance as opposed to being aspontaneous reaction to a person’s perception of the immediate situation. That is, the indi-vidual may reflect and deliberate about a behavioral plan and may decide how he or sheintends to behave. In so doing, the person may consciously consider the implications of hisor her attitude. For example, when buying a car or deciding which college to attend, a personwill extensively deliberate about the decision and consider all of the advantages and disad-vantages before making a behavioral decision. Alternatively, the individual might not activelyreflect on his or her attitude, but that attitude may influence how the person interprets theevent that is occurring and, in that way, may affect the behavior. When choosing between apistachio ice cream cone and a chocolate one, a person will rarely analyze the positive and

Acting as We Feel— 51

03-Brock.qxd 12/1/2004 10:51 AM Page 51

negative features of each flavor. Instead, the individual’s attitudes toward the different flavorsdetermine which flavor looks better at that moment in time. The former type of process is theessence of Ajzen and Fishbein’s (1980) theory of reasoned action. The latter is depicted inFazio’s (1986) model of the attitude-to-behavior process.

Ajzen and Fishbein’s Theory of Reasoned Action

As implied by its name, the theory of reasoned action assumes that people deliberate aboutthe wisdom of a given course of action:

We argue that people consider the implications of their actions before they decide to engageor not engage in a given behavior. For this reason, we refer to our approach as a “theory ofreasoned action.” . . . We make the assumption that most actions of social relevance areunder volitional control and, consistent with this assumption, our theory views a person’sintention to perform (or to not perform) a behavior as the immediate determinant of action.(Ajzen & Fishbein, 1980, p. 5)

According to this theory, then, an individual’s behavioral intention is the single bestpredictor of his or her eventual behavior. The theory goes on to specify the factors that anindividual considers in forming a behavioral intention. The person considers, weighs, andcombines (a) his or her attitude toward the behavior in question and (b) subjective normsregarding the behavior (Figure 3.3). The second component, subjective norms, involves boththe person’s beliefs about what important others think he or she should do and the person’smotivation to comply with the wishes of these others. In deciding whether to attend college,an individual may consider what his or her friends and parents think about attending collegeas well as how important it is to comply with the wishes of his or her friends and parents.

52— P E R S U A S I O N

The person’s beliefs that specificindividuals or groups think heshould or should not perform thebehavior and his motivation tocomply with the specific referents

The person’s beliefs that the behavior leads to certainoutcomes and his evaluationsof those outcomes

Attitude towardthe behavior

Subjectivenorm

Intention Behavior

Figure 3.3 Schematic Diagram of Ajzen and Fishbein’s Theory of Reasoned Action

SOURCE: From Ajzen, I., & Fishbein, M., Understanding Attitudes and Predicting Social Behavior. Copyright ©1980. Reprinted with permission of Prentice-Hall.

03-Brock.qxd 12/1/2004 10:51 AM Page 52

The first component, the attitude component, refers specifically to the behavioral choiceunder consideration (e.g., buying a 50-inch plasma high-definition television), not the generalattitude toward the object (e.g., a 50-inch plasma high-definition television). The reason whythe model differentiates between attitudes toward behavior and attitudes toward the object isthat a person may have a positive attitude toward a 50-inch plasma high-definition televisionbut not toward buying it due to the costs involved in such a behavior (e.g., increased creditpayments, finding a place to put it). According to the theory, the individual constructs this atti-tude toward the behavior by a careful analysis of available information. The attitude is a func-tion of the person’s beliefs concerning the likely outcomes to result from performing thebehavior and the person’s positive or negative feelings about those outcomes.

As an example, consider a young couple who are deciding whether to have a baby.According to the theory of reasoned action, the couple would consider the outcomes that arelikely to occur if they were to have a baby (e.g., having to nurture the baby, having less timeto engage in leisure activities, playing with the baby, facing strain on the family budget) andtheir evaluations of these outcomes. From this information, they would construct an attitudetoward having a baby. The couple would also consider how people who are important to them(i.e., family and friends) would feel about their having a child. Are their family and friendspressuring the couple to have a child? Would such significant others be supportive? The morepositively the couple view the prospect of having a baby and the more support they perceivefrom others for their doing so, the more likely the couple will arrive at the intention to havea child.

In a study of family planning, Davidson and Jaccard (1979) found among a sample ofwomen a strong correlation between behavioral intentions to have children and their actualchildbearing over the next 2 years. Furthermore, the women’s behavioral intentions werehighly related to measures of the attitudinal and normative components specified by thetheory. This investigation, along with a number of other studies reviewed by Ajzen andFishbein (1980), provides impressive support for the theory of reasoned action.

Fazio’s Attitude-to-Behavior Process Model

The theory of reasoned action assumes that attitudes guide behavior through consciousconsideration of and deliberation about a person’s attitude and its implications for a givencourse of action. In contrast, the process model proposed by Fazio and colleagues (Fazio,1986; Fazio, Powell, & Herr, 1983) suggests that attitudes can guide a person’s behavior evenwhen the person does not actively reflect and deliberate about the attitude. When someonesees a cockroach, he or she probably does not consider the beliefs about how unsanitary cock-roaches are, nor is the person likely to reason about what other people think of smashing thecockroach. If people did engage in such extensive thinking, the cockroach would disappearbefore anyone had a chance to decide how to react. Instead, in very basic terms, the processmodel argues that the individual’s attitude toward cockroaches would define this situation asan unpleasant one and that the person would act on this feeling or impulse.

According to the process model, the precursor of behavior is an individual’s definition ofthe event that is occurring. That is, the individual’s interpretation of what is happening isassumed to determine how he or she responds. Classic research concerning bystander inter-vention provides a useful illustration of this principle (Latane & Darley, 1970). In one exper-iment, participants were led into a room and asked to fill out a questionnaire while the

Acting as We Feel— 53

03-Brock.qxd 12/1/2004 10:51 AM Page 53

experimenter waited outside the room. A short time later, smoke-like vapors started to enterthe room through a vent in the wall. Some participants perceived the vapors as harmless,whereas others interpreted them as actual smoke from a fire. Defining the event as a fire wasfound to be a critical step if an individual was to react to the emergency situation by notifyingthe experimenter. In this “smoke-filled room” experiment, people who failed to define thesmoke-like vapors as an indication of a fire were unlikely to report the incident, even thoughthe vapors eventually became so dense that individuals had difficulty in reading their ques-tionnaires. Thus, how people perceive a situation has a profound impact on their behavior.The same set of people can be taken to the same physical setting, and they may respond verydifferently.

Within the process model, this definition of the event consists of two components: an indi-vidual’s perceptions of the attitude object in the immediate situation and the individual’s def-inition of the situation. Definition of the situation refers to the storehouse of knowledge thatthe individual possesses concerning behaviors that are to be expected and that are appropri-ate in the particular situation. For example, when smoke enters a room due to a fire, the normis that people should report the fire. It is in this way that norms can influence behavior.

Perceptions of the attitude object in the immediate situation also influence people’s defin-ition of an event and provide the means for a potential impact of attitudes. A vast literatureindicates that attitudes can guide how and what people perceive (for a review, see Fazio,1986). In the words of Smith, Bruner, and White (1956), an attitude provides “a ready aid in‘sizing up’ objects and events in the environment” (p. 41) (Figure 3.4).

Earlier we discussed the notion that attitudes can vary in how accessible they are from mem-ory. The idea that attitudes vary in their accessibility from memory is an integral aspect of theprocess by which attitudes can influence what a person perceives. The Fazio process model viewsan attitude as an association in memory between the attitude object and a person’s evaluation ofthe object. The strength of this association can vary and determines the accessibility of the atti-tude from memory. To illustrate, consider an example outside the context of attitudes—the asso-ciation between bacon and eggs. If someone mentions bacon to you, you probably cannot helpbut think of eggs because the two are so strongly associated. On the other hand, if someone says“sidewalk,” you are much less likely to think of eggs despite the old saying that sometimes it is

54— P E R S U A S I O N

Figure 3.4 Illustration of How Attitudes Guide Our Perceptions and Judgments

SOURCE: Calvin and Hobbes © 1987 Watterson. Reprinted with permission of Universal Press Syndicate. All rightsreserved.

03-Brock.qxd 12/1/2004 10:51 AM Page 54

so hot you can fry an egg on the sidewalk. The bacon–eggs association is much stronger inmemory than is the sidewalk–eggs association. Likewise, the strength of the association betweenan attitude object and a person’s evaluation of the object can vary. As discussed earlier, theassociation (e.g., between cockroach and “Yuck!”) can be so strong that the evaluation comes tomind immediately and spontaneously when a person encounters the attitude object.

The process model maintains that the attitude must be activated from memory if the atti-tude is to exert any influence over a person’s behavior. If activated, the attitude acts as a fil-ter through which the object is viewed at that moment in time. As a result, immediateperceptions of the attitude object will be consistent with the attitude. In contrast, if the atti-tude is not activated, the immediate perceptions will be based on momentarily noticeable fea-tures of the attitude object that might not be consistent with the attitude. For example, whennoticing a grocery item toward which you do not have an accessible attitude, features such asthe type of wrapping, the position of the item on the shelf, and whether the item is on sale arelikely to influence your immediate perception of the item.

According to the process model, then, the initiation of the attitude-to-behavior processdepends on whether the attitude is activated from memory. There are a number of ways inwhich attitudes can be activated from memory. Such activation can occur as a result ofsituational cues (Snyder & Kendzierski, 1982). When we are told to vote our feelings, our atti-tudes are likely to be activated from memory. However, attitude activation also can occurwithout the benefit of prompting from a situational cue if the attitude is sufficiently accessi-ble from memory. As we discussed earlier, attitudes involving strong object–evaluation asso-ciations (e.g., cockroach and “Yuck!”) are highly accessible from memory and can beactivated from memory automatically or effortlessly merely on seeing the attitude object. It issuch attitudes that are capable of initiating the attitude–behavior process, even withoutprompting from any situational cue. Once activated from memory, the attitude can influencea person’s perceptions of the object in the immediate situation, his or her definition of theevent, and (ultimately) his or her behavior (Figure 3.5).2

The model predicts that attitude accessibility will determine the relation between attitudesand perceptions or judgments of an object. The relation is expected to be stronger if theattitude is accessible from memory than if it is not. The model makes a similar prediction

Acting as We Feel— 55

Immediateperceptions ofthe attitudeobject

Definitionof thesituation

Norms

Definitionof theevent

BehaviorSelectiveperception

Attitudeactivation

Figure 3.5 Schematic Diagram of Fazio’s Attitude-to-Behavior Process Model

SOURCE: From Fazio, R., “How do attitudes guide behavior?” in The Handbook of Motivation and Cognition:Foundations of Social Behavior, edited by Sorrentino, R. M., & Higgins, E. T. Copyright © 1986. Reprinted withpermission of Guilford Press.

03-Brock.qxd 12/1/2004 10:51 AM Page 55

regarding the attitude–behavior relation. Both of these hypotheses have been supported inresearch (for a full review of this research, see Fazio, 1995). In a study of the relation betweenattitudes toward former President Ronald Reagan and judgments of the performance of can-didates in the televised debates during the 1984 election campaign, the attitudes of a sampleof townspeople were assessed along with the accessibility of those attitudes as indicated byhow long people took to respond to an attitude question (Fazio & Williams, 1986). Followingthe debates, the respondents were mailed postcards requesting them to indicate which candi-date had performed better in the debates. As one would expect, these judgments were biasedby people’s attitudes. The more positive the attitude toward Reagan, the more likely theperson was to judge Reagan and the Republican vice presidential candidate, George Bush, tohave performed better than their opponents. More important, as predicted by the model, thisbias conferred by one’s attitude was stronger among respondents whose attitudes towardReagan were relatively more accessible from memory than among those with less accessibleattitudes.

This same study also found evidence of the predicted impact of attitude accessibility onthe attitude–behavior relation. After the election, the respondents were telephoned and askedhow they had voted. The relation between attitudes toward Reagan (as measured during mid-summer) and voting was much higher among individuals with accessible attitudes thanamong those with less accessible attitudes.

More evidence of the influence of attitude accessibility on behavior is provided in a studywhere college students’ attitudes and the accessibility of the attitude toward a variety of prod-ucts were measured (Fazio, Powell, & Williams, 1989). As a token of appreciation for havingparticipated in the study, at the end of the study the students were shown 10 products andallowed to select 5 of them to take home. The available products included items such asSnickers and Mounds candy bars, Sun-Maid raisins, Planters peanuts, and V8 juice. The prod-ucts that the students selected served as the behavior measure. As predicted by the processmodel, the relation between a student’s attitude toward a given product and whether thestudent chose the given product was found to increase as the accessibility of the attitudeincreased. For example, a student who reported liking Planters peanuts but who took a rela-tively long time to report that attitude was less likely to choose Planters as his or her free giftthan was another student who reported liking Planters peanuts equally well but reported thatattitude faster.

According to the process model, then, whether an attitude directs behavior will depend onwhether it is activated from memory and the extent to which it colors individuals’ definitionof the event. Through such a process, attitudes can serve as remarkably functional tools forindividuals. Attitudes that are accessible from memory can guide individuals’ behavior ina satisfying direction without the individuals having to engage in conscious deliberativereasoning. In this way, they can simplify day-to-day life.

The MODE Model: An Integration

Two different processes by which attitudes can guide behavior have been discussed. Oneprocess clearly focuses on deliberate, planned, and reasoned action in which an attitude exertsan impact on behavior because an individual reflects on the attitude. The other process con-cerns an influence of attitude that need not stem from reflection but instead can stem from theattitude’s influence on the person’s perception of objects and situations that, in turn, affects

56— P E R S U A S I O N

03-Brock.qxd 12/1/2004 10:51 AM Page 56

his or her behavior. Clearly, not all social behavior is deliberate and reasoned. Just as clearly,not all behavior is an impulsive reaction to a person’s definition of the event.

Given that both processes occur, under what conditions is each process likely to operate?The MODE (Motivation and Opportunity and DEterminants) model integrates these twoprocess models within a single framework. In brief, we would argue that which processoccurs depends on both motivation and opportunity (Fazio, 1990; Fazio & Towles-Schwen,1999). The deliberate process proposed by the theory of reasoned action obviously requiresextensive cognitive work. As a result, it seems reasonable to propose that some motivation isnecessary to induce individuals to engage in this effortful reasoning process. Such motivationis likely to exist when an individual’s behavioral decision will have important consequences.When the cost of making a bad decision is perceived to be high, the individual will be moti-vated to engage in careful reasoning (Kruglanski & Freund, 1983). Without such inducement,the individuals might see little cost to permitting behavior to flow spontaneously from his orher interpretation of the event. Of course, the motivation to engage in effortful reasoning isnot sufficient in and of itself. The opportunity to do so also must exist. Situations that requirethe individual to make a behavioral response quickly can deny him or her the chance toengage in much reasoning. Thus, the MODE model maintains that the process depicted by thetheory of reasoned action is more likely to occur when the situation both motivates the indi-vidual to consider his or her action carefully and allows the individual the opportunity to doso. Without such motivation or opportunity, the more spontaneous process proposed by Fazio(1986) may be more likely.

In a test of the MODE model, Sanbonmatsu and Fazio (1990) gave participants informationabout two different department stores. The description of the stores was constructed such thatone of the stores (Smith’s) was clearly better overall than the other store (Brown’s). However,Brown’s was described by statements indicating that it had the better camera department. In thecritical task, participants were asked to judge which store they would visit to purchase a cam-era. The aim of the experiment was to discern the conditions under which participants wouldundertake the effort to search their memories for the detailed information they had receivedabout the two stores’ respective camera departments as opposed to simply relying on their senseof which store was generally better. Participants’ motivation to make a correct decision wasmanipulated by telling some of them that they would have to justify their decision to a group ofundergraduate students and the experimenter (high motivation). The remaining participantswere simply asked to choose which store they would visit to buy the camera (low motivation).In addition, some participants were forced to make the decision quickly (high time pressure),whereas other participants could make the decision at their own pace (low time pressure).

According to the MODE model, when participants are highly motivated and have theopportunity to carefully deliberate, they should consider the information they had receivedextensively while making a choice. In this instance, participants should choose Brown’s overSmith’s, despite the fact that Smith’s is the better overall store, because Brown’s has the bettercamera department. However, when participants are not motivated and/or do not have theopportunity to deliberate, they should make the decision based on their global attitude ratherthan considering the specific information they had received about the two stores’ cameradepartments. In these instances, the participants should choose Smith’s because they have amore positive attitude toward it.

The results were consistent with the MODE model’s predictions. When participants weremotivated and had the opportunity to make a choice, they chose Brown’s because it had the better

Acting as We Feel— 57

03-Brock.qxd 12/1/2004 10:51 AM Page 57

camera department. Only when both motivation and opportunity were high did participantsengage in the effort necessary to retrieve specific details from memory. When participants wereeither less motivated or under time pressure, they decided on the basis of their overall attitude andchose Smith’s.

The research on the MODE model demonstrates that when people are making spontaneousdecisions, accessible attitudes will be more predictive of their behavior. However, whenpeople are making deliberate decisions, accessible attitudes will not necessarily be as predic-tive of their behavior. Interestingly, recent research has demonstrated that accessible attitudescan also influence whether people make more deliberate decisions as outlined by the theoryof reasoned action. The attitude-as-information hypothesis proposes that accessible attitudescan act as cues that something important is in the environment (Roskos-Ewoldsen, Bischell,& Hoffman, 2002). The basic idea is that the attitudinal reaction that people experience whenan accessible attitude is activated signals that something of importance is happening. This cuewill motivate people to more carefully deliberate on the available information in the environ-ment (Fabrigar, Priester, Petty, & Wegener, 1998).

By cuing the importance, accessible attitudes also may motivate people to engage in thetype of thoughtful processes outlined by the theory of reasoned action. For example, in a recentexperiment, women were presented with messages concerning breast cancer and breast self-exams (Roskos-Ewoldsen, Yu, & Rhodes, 2004). Half of the messages stressed the dangers ofbreast cancer (high fear), and the other half downplayed the risks of breast cancer (low fear).In addition, half of the messages stressed the effectiveness of breast self-exams for the earlydetection and treatment of breast cancer (high efficacy), and the other half focused on the lim-its of breast self-exams (low efficacy). A short time after hearing the messages, the accessibil-ity of the women’s attitudes to breast cancer and breast self-exams was measured, as were theirintentions to perform breast self-exams in the future. Women who heard the high efficacy mes-sages, regardless of the level of fear, developed more accessible attitudes toward breast self-exams. In addition, as the accessibility of the women’s attitudes toward breast self-examsincreased, their intention to perform breast self-exams in the future increased. Of course,behavioral intentions are a key part of the theory of reasoned action, and this finding suggeststhat the accessible attitudes toward breast self-exams motivated these women to carefullydeliberate about the self-exams and come to a deliberate decision to perform them in the future.

CONCLUSIONS

The pessimism concerning the relationship between attitudes and behavior during the 1970swas clearly unwarranted. Attitudes can predict behavior. As we have seen in this chapter, anumber of factors influence when attitudes predict behavior, including characteristics of thebehavior, the attitudes, the situation, and the person. More important, psychological modelsof the attitude–behavior relationship have been developed, including the theory of reasonedaction (Fishbein & Ajzen, 1975) and Fazio’s (1986) process model. More recently, the MODEmodel (Fazio, 1990) has sought to combine the theory of reasoned action and the processmodel within a single framework. The refinements of the MODE model demonstrate how farour understanding of the attitude–behavior relationship has come since LaPiere’s (1934) clas-sic research. The MODE model also demonstrates the complexities of understanding whenand how attitudes will predict behavior.

58— P E R S U A S I O N

03-Brock.qxd 12/1/2004 10:51 AM Page 58

NOTES

1. LaPiere’s (1934) study plays another important role in history besides serving as a classicexample of the failure of attitudes to predict behaviors. The U.S. Supreme Court relied in part onLaPiere’s and subsequent research in the landmark Brown v. Board of Education decision that endedsegregation in education in the United States. LaPiere’s study focused on how people acted when, in thiscase, Chinese individuals integrated a segregated establishment. No violence was aimed at LaPiere andthe Chinese couple who were traveling with him when the establishments had been segregated againstChinese. Other studies also found that people did not react violently when blacks integrated segregatedstores. These studies were used to demonstrate to the Supreme Court that ending segregation would notnecessarily result in violence (Jackson, 2001).

2. Research findings concerning attitude accessibility have led to two conclusions regarding itsdeterminants and assessment. First, the accessibility of an attitude and the likelihood that the attitudewill be activated from memory on mere observation of the attitude object depends on the strength of theobject–evaluation association (Fazio, Chen, McDonel, & Sherman, 1982; Fazio, Powell, & Herr, 1983;Fazio, Sanbonmatsu, Powell, & Kardes, 1986). Second, the strength of the object–evaluation associa-tion can be reasonably estimated by measuring how quickly an individual can respond to an inquiryabout his or her attitude toward the object (Fazio, 1993; Fazio et al., 1986). Relatively fast responsesindicate strong associations and relatively high accessibility from memory.

REFERENCES

Ajzen, I., Darroch, R., Fishbein, M., & Hornik, J. (1970). Looking backward revisited: A reply toDeutscher. American Sociologist, 5, 267–273.

Ajzen, I., & Fishbein, M. (1977). Attitude–behavior relations: A theoretical analysis and review ofempirical research. Psychological Bulletin, 84, 888–918.

Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behavior. EnglewoodCliffs, NJ: Prentice Hall.

Ajzen, I., Timko, C., & White, J. B. (1982). Self-monitoring and the attitude–behavior relation. Journalof Personality and Social Psychology, 42, 426–435.

Amodio, D. M., & Devine, P. G. (2005). Changing prejudice: The effects of persuasion on implicit andexplicit forms of race bias. In T. C. Brock & M. C. Green (Eds.), Persuasion: Psychologicalinsights and perspectives (2nd ed., chap. 11). Thousand Oaks, CA: Sage.

Bechtold, A., Naccarato, M. E., & Zanna, M. P. (1986). Need for structure and the prejudice–discriminationlink. Paper presented at the annual meeting of the Canadian Psychological Association, Toronto.

Blascovich, J., Ernst, J. M., Tomaka, J., Kelsey, R. M., Salomon, K. L., & Fazio, R. H. (1993). Attitudeaccessibility as a moderator of autonomic reactivity during decision making. Journal of Personalityand Social Psychology, 64, 165–176.

Calder, B. J., & Ross, M. (1973). Attitudes and behavior. Morristown, NJ: General Learning Press.Corey, S. M. (1937). Professed attitudes and actual behavior. Journal of Educational Psychology, 28,

271–280.Davidson, A. R., & Jaccard, J. J. (1979). Variables that moderate the attitude–behavior relation: Results

of a longitudinal survey. Journal of Personality and Social Psychology, 37, 1364–1376.Deutscher, I. (1973). Why do they say one thing, do another? Morristown, NJ: General Learning Press.Dillehay, R. C. (1973). On the irrelevance of the classical negative evidence concerning the effect of

attitudes on behavior. American Psychologist, 28, 887–891.Fabrigar, L. R., Priester, J. R., Petty, R. E., & Wegener, D. T. (1998). The impact of attitude accessibil-

ity on elaboration of persuasive messages. Personality and Social Psychology Bulletin, 24,339–352.

Acting as We Feel— 59

03-Brock.qxd 12/1/2004 10:51 AM Page 59

Fazio, R. H. (1986). How do attitudes guide behavior? In R. M. Sorrentino & E. T. Higgins (Eds.), Thehandbook of motivation and cognition: Foundations of social behavior (pp. 204–243). New York:Guilford.

Fazio, R. H. (1990). Multiple processes by which attitudes guide behavior: The MODE model as anintegrative framework. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 23,pp. 75–109). San Diego, CA: Academic Press.

Fazio, R. H. (1993). Variability in the likelihood of automatic attitude activation: Data re-analysis andcommentary on Bargh, Chaiken, Govender, and Pratto (1992). Journal of Personality and SocialPsychology, 64, 753–758, 764–765.

Fazio, R. H. (1995). Attitudes as object–evaluation associations: Determinants, consequences, andcorrelates of attitude accessibility. In R. E. Petty & J. A. Krosnick (Eds.), Attitude strength:Antecedents and consequences (pp. 247–282). Mahwah, NJ: Lawrence Erlbaum.

Fazio, R. H. (2000). Accessible attitudes as tools for object appraisal: Their costs and benefits. InG. Maio & J. Olson (Eds.), Why we evaluate: Functions of attitudes (pp. 1–36). Mahwah, NJ:Lawrence Erlbaum.

Fazio, R. H., Blascovich, J., & Driscoll, D. M. (1992). On the functional value of attitudes: The influ-ence of accessible attitudes upon the ease and quality of decision making. Personality and SocialPsychology Bulletin, 18, 388–401.

Fazio, R. H., Chen, J., McDonel, E. C., & Sherman, S. J. (1982). Attitude accessibility, attitude–behaviorconsistency, and the strength of the object–evaluation association. Journal of Experimental SocialPsychology, 18, 339–357.

Fazio, R. H., & Dunton, B. C. (1997). Categorization by race: The impact of automatic and controlledcomponents of racial prejudice. Journal of Experimental Social Psychology, 33, 451–470.

Fazio, R. H., Ledbetter, J. E., & Towles-Schwen, T. (2000). On the costs of accessible attitudes:Detecting that the attitude object has changed. Journal of Personality and Social Psychology, 78,197–210.

Fazio, R. H., & Powell, M. C. (1997). On the value of knowing one’s likes and dislikes: Attitude acces-sibility, stress, and health in college. Psychological Science, 8, 430–436.

Fazio, R. H., Powell, M. C., & Herr, P. M. (1983). Toward a process model of the attitude–behavior rela-tion: Accessing one’s attitude upon mere observation of the attitude object. Journal of Personalityand Social Psychology, 44, 723–735.

Fazio, R. H., Powell, M. C., & Williams, C. J. (1989). The role of attitude accessibility in the attitude-to-behavior process. Journal of Consumer Research, 16, 280–288.

Fazio, R. H., Sanbonmatsu, D. M., Powell, M. C., & Kardes, F. R. (1986). On the automatic activationof attitudes. Journal of Personality and Social Psychology, 50, 229–238.

Fazio, R. H., & Towles-Schwen, T. (1999). The MODE model of attitude–behavior processes. InS. Chaiken & Y. Trope (Eds.), Dual process theories in social psychology (pp. 97–116). New York:Guilford.

Fazio, R. H., & Williams, C. J. (1986). Attitude accessibility as a moderator of the attitude–perceptionand attitude–behavior relations: An investigation of the 1984 presidential election. Journal ofPersonality and Social Psychology, 51, 505–514.

Fazio, R. H., & Zanna, M. P. (1978). Attitudinal qualities relating to the strength of the attitude–behaviorrelationship. Journal of Experimental Social Psychology, 14, 398–408.

Fazio, R. H., & Zanna, M. P. (1981). Direct experience and attitude–behavior consistency. InL. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 14, pp. 162–202).New York: Academic Press.

Fishbein, M., & Ajzen, I. (1974). Attitudes toward objects as predictors of single and multiple behav-ioral criteria. Psychological Review, 81, 59–74.

Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction to theory andresearch. Reading, MA: Addison–Wesley.

60— P E R S U A S I O N

03-Brock.qxd 12/1/2004 10:51 AM Page 60

Goodmonson, C., & Glaudin, V. (1971). The relationship of commitment-free behavior and commitmentbehavior: A study of attitude toward organ transplantation. Journal of Social Issues, 27, 171–183.

Heberlein, T. A., & Black, J. S. (1976). Attitudinal specificity and the prediction of behavior in a fieldsetting. Journal of Personality and Social Psychology, 33, 474–479.

Jackson, J. P., Jr. (2001). Social scientists for social justice: Making the case against segregation.New York: New York University Press.

Jamieson, D. W., & Zanna, M. P. (1985). Moderating the attitude–behavior relation: The joint effects ofarousal and self-monitoring. Paper presented at the annual meeting of the Canadian PsychologicalAssociation, Halifax, Nova Scotia.

Jamieson, D. W., & Zanna, M. P. (1989). Need for structure in attitude formation and expression. InA. R. Pratkanis, S. J. Breckler, & A. G. Greenwald (Eds.), Attitude structure and function(pp. 383–406). Hillsdale, NJ: Lawrence Erlbaum.

Kelley, S., & Mirer, T. W. (1974). The simple act of voting. American Political Science Review, 68, 572–591.Kruglanski, A. W., & Freund, T. (1983). The freezing and unfreezing of lay-inferences: Effects on

impressional primacy, ethnic stereotypic, and numerical anchoring. Journal of Experimental SocialPsychology, 19, 448–468.

Kutner, B., Wilkins, C., & Yarrow, P. R. (1952). Verbal attitudes and overt behavior involving racial prej-udice. Journal of Abnormal and Social Psychology, 47, 649–652.

LaPiere, R. T. (1934). Attitudes vs. actions. Social Forces, 13, 230–237.Latane, B., & Darley, J. M. (1970). The unresponsive bystander: Why doesn’t he help? New York:

Appleton–Century–Crofts.Lord, C. G., Lepper, M. R., & Mackie, D. (1984). Attitude prototypes as determinants of attitude–behavior

consistency. Journal of Personality and Social Psychology, 46, 1245–1266.Petty, R. E., & Krosnick, J. A. (Eds.). (1995). Attitude strength: Antecedents and consequences.

Mahwah, NJ: Lawrence Erlbaum.Regan, D. T., & Fazio, R. H. (1977). On the consistency between attitudes and behavior: Look to the

method of attitude formation. Journal of Experimental Social Psychology, 13, 38–45.Rholes, W. S., & Bailey, S. (1983). The effects of level of moral reasoning on consistency between moral

attitudes and related behaviors. Social Cognition, 2, 32–48.Roskos-Ewoldsen, D. R. (1997). Attitude accessibility and persuasion: Review and a transactive model.

In B. Burleson (Ed.), Communication yearbook 20 (pp. 185–225). Thousand Oaks, CA: Sage.Roskos-Ewoldsen, D. R., Bichsel, J., & Hoffman, K. (2002). The influence of accessibility of source

likability on persuasion. Journal of Experimental Social Psychology, 38, 137–143.Roskos-Ewoldsen, D. R., & Fazio, R. H. (1992). On the orienting value of attitudes: Attitude accessi-

bility as a determinant of an object’s attraction of visual attention. Journal of Personality andSocial Psychology, 63, 198–211.

Roskos-Ewoldsen, D., Yu, H. J., & Rhodes, N. (2004). Fear appeal messages affect accessibility ofattitudes toward the threat and adaptive behaviors. Communication Monographs, 71, 49–69.

Sanbonmatsu, D. M., & Fazio, R. H. (1990). The role of attitudes in memory-based decision making.Journal of Personality and Social Psychology, 59, 614–622.

Schuman, H., & Johnson, M. P. (1976). Attitudes and behavior. Annual Review of Sociology, 2, 161–207.Seligman, C., Kriss, M., Darley, J. M., Fazio, R. H., Becker, L. J., & Pryor, J. B. (1979). Predicting summer

energy consumption from home-owners’ attitudes. Journal of Applied Social Psychology, 9, 70–90.Smith, E. R., Fazio, R. H., & Cejka, M. A. (1996). Accessible attitudes influence categorization of

multiply categorizable objects. Journal of Personality and Social Psychology, 71, 888–898.Smith, F. J. (1977). Work attitudes as predictors of attendance on a specific day. Journal of Applied

Psychology, 62, 16–19.Smith, M. B., Bruner, J. S., & White, R. W. (1956). Opinions and personality. New York: John Wiley.Snyder, M. (1987). Public appearances/Private realities: The psychology of self-monitoring. New York:

Freeman.

Acting as We Feel— 61

03-Brock.qxd 12/1/2004 10:51 AM Page 61

Snyder, M., & Kendzierski, D. (1982). Acting on one’s attitude: Procedures for linking attitude andbehavior. Journal of Experimental Social Psychology, 18, 165–183.

Snyder, M., & Swann, W. B., Jr. (1976). When actions reflect attitudes: The politics of impressionmanagement. Journal of Personality and Social Psychology, 34, 1034–1042.

Vroom, V. H. (1962). Ego-involvement, job satisfaction, and job performance. Personnel Psychology,15, 159–177.

Wicker, A. W. (1969). Attitudes versus actions: The relationship of verbal and overt behavioralresponses to attitude objects. Journal of Social Issues, 25, 41–78.

Wicker, A. W. (1971). An examination of the “other variable” explanation of attitude–behavior incon-sistency. Journal of Personality and Social Psychology, 19, 18–30.

Zanna, M. P., & Fazio, R. H. (1982). The attitude–behavior relation: Moving toward a third generationof research. In M. P. Zanna, E. T. Higgins, & C. P. Herman (Eds.), Consistency in social behavior:The Ontario Symposium (Vol. 2, pp. 283–301). Hillsdale, NJ: Lawrence Erlbaum.

Zanna, M. P., Olson, J. M., & Fazio, R. H. (1980). Attitude–behavior consistency: An individual differ-ence perspective. Journal of Personality and Social Psychology, 38, 432–440.

62— P E R S U A S I O N

03-Brock.qxd 12/1/2004 10:51 AM Page 62